Old Habits and Preconceived Notions
The Behringer Ultra-Drive Pro DCX2496 that comes with the Models 10’s, is a digital crossover, and inserts immediately after the preamp. Thus, analog signals that originate from a DAC or phono stage, or tape player, are sent to the preamp, thence to the Behringer where they are (1) converted to digital by the Behringer’s A-to-D converter, (2) processed in the digital domain, then (3) converted back to analog in the Behringer’s D-to-A converter. The Behringer retails for about $400, whereas my AMR DP-777 DAC retails for $5,000. Despite how good the Model 10’s always sounded at shows, it didn’t seem possible that the Behringer wouldn’t seriously degrade the signal from my AMR — a case of preconceived notions running smack in the face of actual experience. Even worse was the thought of digitizing the signal from my vinyl playback system; this heretic act would surely be ruinous. Earlier versions of the Model 10 used an analog crossover, so when I arranged for the review, I asked Roger if I could try one of those as well, to which he agreed. The analog crossover was built into his ESL amplifier but as that amplifier has been upgraded, I wanted to use the current-production amplifier, with both the digital and analog crossovers. Fortunately, the amp section in the older unit can be bypassed, so I was able to use just the crossover.
It should be noted that the comparison wasn’t purely that of analog to digital, as other differences exist between the two crossovers, for reasons I’ll now explain. The fundamental resonance of the Model 10 electrostat panel is around 86 Hz. The output from the crossover must be down by at least 48 dB at this frequency to assure that the resonance is not excited. As Roger explained, “The resonance will degrade the bass frequency response. The resonance is high Q, which will cause the bass quality to become muddy, loose, flabby, and poorly defined due to a lot of overshoot and ringing. So it is essential to keep the crossover point high enough that the fundamental resonance is not excited.” For his analog crossover, Roger chose a slope of 24 dB/octave. All else being equal, a steeper slope is preferable but, in Roger’s words, “making steeper slopes with passive inductors, resistors, and capacitors becomes very difficult because the tolerance of the components is so poor that you can’t get their poles to match precisely enough to work correctly. After all, most capacitors have a tolerance of +/- 20%, and it is common for large capacitors like those used in speaker crossovers to have tolerances of +/- 50%. Inductors have only slightly better tolerances.”
To achieve the requirement of the signal being down 48 dB at the resonant frequency using a 24 dB slope, the crossover point had to be at 344 Hz or higher. Roger chose a crossover point of 350 Hz for his analog crossover. Unfortunately, this means that the woofer contributes significant energy all the way up to 1,400 Hz, at which frequency its output is down by 48 dB. Thus, the woofer’s output mixes with that of the electrostatic panel throughout most of the midrange, i.e., 500 Hz to 2 KHz. This degrades the clarity and detail of the midrange, which is best reproduced solely by the nearly-massless electrostatic panel.
In contrast to analog crossovers, digital crossovers, such as that in the Behringer, come with selectable crossover slopes and filter types. Roger chose a 48 dB/octave crossover slope, which made “it possible to completely eliminate the contribution of each driver within just one octave of the crossover point, thereby reducing the shared bandwidth to a minimum and greatly reducing the stress on the drivers.” Using a crossover point of 172 Hz, the output of the panel at its fundamental resonance is down by 48 dB, and the woofer output is down by 48 dB at 344 Hz, thus there is essentially no woofer energy in the midrange. Of course, to compare apples to apples, we could have re-programmed the Behringer for a 350 Hz crossover, but that was more work than either of us wanted to undertake.
Yet another difference relates to time alignment. The Behringer, but not the analog crossover, allows one to delay the early-arriving driver, resulting in perfect time-alignment between the electrostat panel and the woofer. This, too, is part of the pre-programming Roger provides.
So while the comparison between analog and digital crossovers was far from perfectly controlled, it was nonetheless quite informative. Upon first listening I preferred the analog crossover; it sounded warmer, and more “musical.” Over the course of a few days, I realized that in fact the sound was muddier, and more veiled, than when the Behringer was in place. This was a “teachable moment” because I realized that after so many years of hearing a veiled sound, it had became normal and “comfortable” to me. In fact, many people have this reaction when first hearing the Model 10’s; the speakers are so incredibly transparent that it seems like something is missing, but that something is coloration and distortion.
But returning to my original concerns of the digital crossover, it was apparent that it was not doing anything terribly wrong. Whether or not an analogue crossover with 48 dB slopes and a 172 Hz crossover point would sound better than the Behringer, remains an open question. Mostly to satisfy my curiosity, I am looking into obtaining and testing such a crossover.
But incredibly, with the Behringer in the system, vinyl still sounds like, well, vinyl. When guests come over I usually begin with digital recordings. Soon as I switch to vinyl they invariably comment on how much better vinyl is than digital. I then point out to them that the signal has in fact being digitized. Not uncommonly, I am greeted with blank stares, and I have to explain — slowly — the precise signal path, and what exactly the Behringer is doing. The blank stares then change to ones of disbelief. I confess to getting a perverse thrill out of this.
But again in the interests of full disclosure, some listeners claim to hear “the cheap Opamps” in the Behringer. I don’t know if they are hearing something that really exists, or they are hearing something that they presume must exist. But truth be told, old habits die hard, and I have not yet completely shaken my own skepticism about the Behringer’s sonics. I am currently investigating alternatives including some that are digital, and will report my findings in a follow-up article. But the take home message is that even if the Behringer is degrading the sound, it must be doing so to a rather small degree, as the resulting sound is world-class.
Conclusions
In the opening paragraph I said that at RMAF I had spoken to Roger about obtaining a pair of Model 10’s for review. That was only a half-truth. In point of fact, what I told Roger was that I wanted to purchase a pair of Model 10’s, and that I would write a review of them. So yes, the “review pair” is, in fact, mine. In the months I have had them, they have re-defined my sense of high-end audio, and brought hour after hour of musical listening pleasure, to both myself and many of my friends and colleagues. My only regret is not having purchased them years ago. The Sanders Sound Model 10’s are amongst the finest speakers I have heard, and certainly the best I have owned. At their price of $14,000 including a stereo amp and active crossover, they are also one of the greatest deals in high-end audio. They unequivocally earn my highest recommendation.
- ← Previous page
- (Page 6 of 6)
I have owned the 10c speakers for a couple of years and continue to enjoy them. I obtained a replacement audio output board for the dcx 2496 that simplifies the circuit, uses better opamps, and also uses a remote control to vary gain…it can be used as a system volume control. A board is also available to replace the digital input / clock circuit. All I can tell you is that I wish Roger would offer this as part of his 10c system, it elevates the transparency, removes grundge, and just makes the speakers better. Contact Ward Maas at Pilghamaudio.com for more info.
Roger does not believe op amps make a difference (he used very inexpensive ones in the Innersound crossover amp, and told a friend of mine who is an op amp guru that “they all sound the same”). I replaced five of the op amps at my friends advice and it made a TREMENDOUS difference. I’m very much a fan of science and like much about objectiveness, but while I was once a “doubter” of some of the tweak nature of high end audio, I’ve come to find over time that numerous things I used to mock I could hear differences in if I gave it a chance and a bit of an open mind. Yes, there is snake oil crap, but there are other things that I have noticed making an improvement (like cables, ha ha ha! Yeah, I used to be in the “all sound the same unless poorly designed” camp, and mocked the cable swapping crowd. Then I stopped mocking and started listening more….)
The op amps I was talking about above (that I replaced) are in the Innersound (now Sanders Sound) crossover amp, not a Behringer unit, just to clarify. Have heard the DCX2496 at shows in Roger’s room, but have not tried it yet in my system).
Thanks for both reviewing a great speaker and interviewing one of the great pioneers in high-end audio. I would like to listen to a pair of Model 10s with a DEQX HDP-4 instead of the Behringer. The DEQX would serve as digital crossover with very steep filters (maybe lowering the crossover point below 172Hz significantly), it would correct both speaker and room, and even seamlessly integrate a pair of subwoofers at any reasonable frequency. Let me just dream.
Glad someone in audio review tells it like it is WRT what live music actually sounds like. Having had subscription series to symphony orchestras and listened to more live jazz than most people, I can tell you that imaging is an pot of gold at the end of the audiophile rainbow. Now, that’s not to say that pinpoint imaging doesn’t have its own visceral enticements, but it’s not what real live music sounds like. So, perhaps we need to abandon the notion that audiophiles must pursue “the closest thing to live”. Maybe we should recognize different tastes and expectations for different audiophiles and leave it at that. Personally, I’ll spend my extra money on live music and more vinyl rather than more exotic gear.
Lots of active crossover options listed here (both analog and digital)
http://thewelltemperedcomputer.com/HW/ActiveCrossover.htm
Has anyone tried a 48dB/oct analog crossover with the model 10s such as the Marchand XM44 and compared with the Behringer?
Chris B
48dB/oct is the standard setting. Roger crosses at 172hz at 48 on both sides.
vdorta
I am test driving a Deqx Premate at the moment, with my 10c s. I am told a HDP4 [ 3 way] has a tweeter x-over that does not go low enough, so you are throwing away a crossover. Save money and buy the 2 way Premate. I have yet to ponder that info. and purchase.
6 weeks later…I have bought the HDP4. For $500 more, you get better transparency and thus imaging. This must be the linear power supply. I saw inside both machines. Anyhow, I thought it no contest in the end. This is so far ahead of my Behringer set up of dcx2496, deq2496 and src2496. I am a happy chappy.
Oh yes…the included asynchronous usb input is truly excellent. I surpasses my Audio-Gd DI V3. I think it injects I2S directly into the digital receiver. I play [a] from a netbook via usb and JRiver,[b] Squeezebox Touch with EDO mod to usb from a SD card [c] cd from Oppo 95 [d] vinyl via SOTA/SME and AQVOX balanced phono stage.
Roger once told me that in order to get the crossover slope that is used in this instance with these speakers, it can only be done in the digital realm to get it low enough. The crossover point in the bass panel amp on my Innersound Eros III amp (his speaker company before Sanders sound, but it’s basically the same speaker), is 310 HZ, higher than what can be done with the DC2496. So that would rule out any analog crossover like the Marchand.
Do you have an active crossover amp, or one of the rare passive crossover Eros models? I have the crossover amp (actually, it’s an Innersound Kaya amp, better transformer and 330 watts at 8 ohms). I have a friend who is an op amp expert, and under his recommendation, I replaced five cheap op amps better (and much more expensive) ones and it made a notable improvement to the sound. I had thought about trying the DCX2496, but afraid the cheap nature of it might ruin the sound (plus I listen to tons of vinyl, have over 3000 lps, so do prefer the idea of not “digitizing” the signal). There are the replacement boards by Pilgham audio mentioned in the first post, so maybe I should pick up a DCX2496 just to try and see if the lower crossover point wins out over my “hotrodded” Innersound crossover amp. If not, I can return the unit. If it shows promise over my crossover amp, then try the upgraded boards mentioned. But….your mention of the Marchand intrigues me. The price is a bit steep (compared to the DCX2496), and I don’t know if Marchand has a return policy, but I do see they sell a board unit only of the crossover which of course is much cheaper. Hmmmm…wonder if it could be installed in my Innersound crossover amp?
I purchased both a Behringer DCX2496 and a Marchand XM44 to try with my Eros MKIII.5. Both tuned to 48dB/oct at 172Hz with the equalization as specified by Roger in his DCX setup manual. They both made the lower midrange much more transparent. There was a “hollow sounding” coloration with the stock crossover. However…. The Behringer obscures microdetail which the Innersound are known for. I preferred the stock crossover for its natural detailed sound despite the lower midrange coloration. Then came the Marchand XM44 with the same tuning as the DCX. The stock Marchand was better than the Behringer for sure but i still felt it was slightly “hifi” sounding. After lots of research, I ordered some dual OPA627 and DEXA discrete opamps from Partsconnexion. I put a pair of DEXA in the input section. I ended up using the OPA627s in the high pass section only since the stock opamps are a little punchier in the low pass. Bingo! This combinations sounds amazing! The lower midrange is seamless now, the bass dynamics and articulation much better, the microdetail is still there, and as an added bonus the high frequencies sound much more extended. Somehow the stock crossover was making the Eros sound muffled. These were always awesome speakers, and now sound even better.
I have the new 10Ds and I have a modded behringer done by Audiosmiles in the UK and it is truly amazing.